Infinity machine creates and checks solutions to open scientific problems
Infinity Machine: a self-improving AI agent for research
Artificial Intelligence
Summary
The Infinity Machine is a computer program designed to solve hard scientific and technical problems by trying out different solutions and checking their correctness automatically. It works by letting smaller AI agents make guesses and then verifying these guesses carefully without human help. The system was able to improve on known results in quantum information science and cut computing energy needs, showing it can make progress on tough questions on its own. However, it sometimes made mistakes or took extra steps when needed. The authors share successes and failures to give a clear picture of what this self-improving AI can do.
What this means in practice
- •For quantum computing developers: Automatically generate and verify solutions to open quantum information problems to improve quantum algorithms and hardware design.
- •For hardware design engineers: Reduce energy consumption and gate count in chip design by automatically optimizing and verifying circuit improvements through a self-improving agent.
Authors
Harish Santhanalakshmi Ganesan
Abstract
We present the Infinity Machine, a self-improving AI agent that we use for research on problems whose answers aprogram or data can check. Each candidate solution is produced by a full tool-using agent that works for up to an hour in a fresh sandbox; proposers see only a public score, and a private split decides what counts. The harness itself(prompts, tools, skills, playbook, meta agent) is an editable genome, changed only through a kernel-enforced gate with held-out and cost checks. An autonomous mode goes from a plain-text goal to a verified report with no human step: an agent picks an open problem and writes its verifier, the kernel calibrates it, an adversarial agent audits it, and an agent that never sees the verifier re-checks the result. Across fourteen problems in six days, for about $180 of model usage, the system produced four results beyond the published state of the art on catalogued open problems in quantum information: a certified depolarizing-channel threshold point p = 0.064966, a qutrit positivity frontier of 0.2910672, a 44-gate depth-6 Golay state preparation, and a stabilizer-rank bound chi(T^7) <= 10. It also cut SABRE qubit-routing overhead from 68,559 to 12,087 added gates. In hardware, it cut the energy per MAC of an open 8x8 INT8 systolic tile by 77.7% through a full open RTL-to-GDS flow, bit-exact. Three fully autonomous runs ended in honestties with public records; one audit caught a stale record, and one agent-written score pointing the wrong way went unnoticed. We report the failures as carefully: the one self-derived harness generation was rejected on cost, proposers gamed lenient scorers four times, and operator changes were decisive for several results.